How to select dataframe values by a list of indices and columns?

Question

I have a dataframe, let's use this example:

df = pd.DataFrame({'A': [5,6,3,4], 'B': [1,2,3,5]})
df
   A  B
0  5  1
1  6  2
2  3  3
3  4  5

and I have a list of columns (by index) that I want to select, let's say:

list=['A','B','B','A']

What I want to obtain is [5,2,3,4], a series if possible. How can I do this? I tried with masks, but I couldn't make it work.

Shubham Sharma · Accepted Answer · 2022-03-31 17:18:15Z

2

You can use lookup:

df.values[range(len(df)), df.columns.get_indexer_for(l)]

array([5, 2, 3, 4], dtype=int64)

answered Mar 31, 2022 at 17:18

Shubham Sharma

71.8k6 gold badges26 silver badges58 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

Upvote for using .values

Haha! Thanks @PanwenWang I always prefer .values over .to_numpy. .values sounds simpler to me :-)

No, I meant using .values or .to_numpy() to avoid looping over df.loc or df.iloc.

@PanwenWang I see.

user7864386 · Accepted Answer · 2022-03-31 17:18:39Z

1

You could create a lookup using numpy advanced indexing:

import numpy as np
idx, cols = pd.factorize(lst)
out = out = df.reindex(cols, axis=1).to_numpy()[range(len(df)), idx].tolist()

Output:

[5, 2, 3, 4]

answered Mar 31, 2022 at 17:18

user7864386